The overall goal of this project is to develop and commercialize a software package to be interfaced with commercial MRI scanners providing physicians with an immediately available diagnostic tool for noninvasive detection, lateralization, and quantification of epileptogenic tissue. This software would convert a set of 2D MRI images of the brain to a 3D percent-pathology map (PPM) with a voxel-by-voxel resolution and offer more patients with epilepsy the prospect of a definitive surgical cure. The techniques implemented and clinically validated in this Phase I proposal would ultimately be developed into the software package and fully validated in a future Phase II proposal. The main target market for this package will be the major manufacturers of MRI scanners. Assessment of pathology is crucial in the pre-surgical evaluation of epilepsy patients, of which, mesial temporal lobe epilepsy (MTLE) patients with hippocampal pathology are the most common. In epilepsy patients with a clear hippocampal lesion such as mesial temporal sclerosis identified by MRI, surgical resection can be curative. Presently available clinical diagnostic MRI, however, fails to identify such a lesion in about 30% of such patients. About half of these "MRI-negative" patients nevertheless do have a distinct pathological lesion discovered post-surgically. Surgery on patients with "MRI negative" epilepsy is associated with poorer post- surgical outcome. The identification of subtle or early abnormalities, presently undetected by MRI but potentially detectable by our method, could help guide the neurosurgeon in selection of tissue for resection. This should lead to improved treatment of patients with epilepsy. The objective of this Phase I proposal is to generate preliminary data that supports the assertion that our method provides the means for noninvasive lateralization and quantification of epileptogenic brain tissue in "MRI-negative" MTLE. This preliminary data would then be used for a future Phase II proposal. This objective will be achieved via these Aims: 1) Generate the PPMs from MRI images in ten MTLE patients with positive conventional diagnostic MRIs (positive controls), ten normal subjects (negative controls), and ten MRI-negative MTLE patients. For comparison to a presently available and commonly used technique, T2 maps will also be generated for each subject. 2) Validate in MTLE patients that PPMs can reliably detect and lateralize epileptogenic hippocampal tissue by comparing the PPMs to the gold standard: intracranial electroencephalography. To show the superiority of our method over presently available MRI methods, head to head comparison will be also made with T2wSI and T2 mapping. 3) Validate in MTLE patients that PPMs yield quantitative information on the degree of hippocampal pathology by comparing our MRI measure of pathology to the gold standard: post-surgical histopathological examination of resected hippocampal tissue. [unreadable] [unreadable] [unreadable]